• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 48 subscribers
  • Views 1470 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

INA331: Startup/Enable Behavior and Large Output Transients

EDY19
Prodigy 40 points
Part Number: INA331


I'm currently designing a load cell front end circuit for a battery powered device.  The load cell is enabled and disabled by a P channel FET, which also enables the INA331 and drives a shunt reference for the INA331.  The goal is to turn on the load cell, enable the reference, and enable the op amp all at approximately the same time so that I can wait for the output to stabilize, take a measurement with a differential ADC, and then shut the load cell and reference back off and disable the op amp.

I'm seeing transients that I can't fully explain when the set of three devices is enabled.  It looks like the amplifier is saturating, taking approximately 150us, and finally recovering and returning to expected behavior.  I've done lots of simulations, and have found that I can replicate the result in simulations by adding imbalance to the capacitors on the load cell lines so that the amplifier saturates.  I also tried adding this imbalance to the hardware, but it didn't seem to affect the amplifier output. 

Next, I tried having the reference always powered, so that only the load cell and amplifier enable were switched on and off - this changed the waveform, but didn't make it any better.  Finally, I kept the load cell and reference enabled so that only the amplifier enable was switched, and the output was stable at the appropriate value. 

The transients aren't an issue in the application, since I can just wait them out before reading the ADC, but I would like to understand them better so that I can have better confidence in the design.

The image below shows the output of the op amp (orange) and the reference voltage (pink).  

 

  • 0
    Guru 140200 points

    Hi Edy,

    I don't think that it has to do with the shutdown feature of INA331. But it is an absolutely typical behaviour of almost all OPAmps that the complex internal circuitry is in imbalance during power-up and that it needs a certain time to get all the individual internal operating points properly stabilized.

    Kai

  • 0 in reply to kai klaas69
    Prodigy 40 points

    I completely agree that most opamps have an undefined output at startup until the internal biases have stabilized, however in this application - the op amp is always powered by a constant 2.8V rail.  It's only the reference, positive, and negative inputs that are switched on and off along with the shutdown pin.  I'm guessing that there are some combinations of common mode input voltage and reference voltage where the op amp isn't happy - and when they pass through the range of ~0-1.2V, the op amp saturates at some point and takes a while to recover.  As I mentioned before, it isn't a huge deal, as the average current drawn by the load cell amplifier has been reduced from ~4mA to <40uA just by changing the on time of the load cell - which made a huge impact on the battery life already!  I just want to be sure that the timing is relatively consistent and isn't hugely affected by component tolerances in an unpredictable manner.

    Thanks!

    Eric

  • 0 in reply to EDY19
    TI__Guru 68541 points

    Eric,

    1.2V is a badgap voltage required for the internal circuitry to begin to settle.  Below such level, the PTAT reference current is not working at all resulting in a wild variation of the internal nodes.  All in all, you should not expect any predictable behaviour until the minimum specified supply voltage is applied (in case of INA331, min Vcc=2.7V).